Documente Academic
Documente Profesional
Documente Cultură
DigitalCommons@UConn
Master's Theses
6-25-2014
Recommended Citation
Barasz, Jonah A., "Experimental Peri-Implant Mucositis in Humans. A Pilot Study" (2014). Master's Theses. Paper 622.
http://digitalcommons.uconn.edu/gs_theses/622
This work is brought to you for free and open access by the University of Connecticut Graduate School at DigitalCommons@UConn. It has been
accepted for inclusion in Master's Theses by an authorized administrator of DigitalCommons@UConn. For more information, please contact
digitalcommons@uconn.edu.
A Thesis
Submitted in Partial Fulfillment of the
Requirements for the Degree of
Master of Dental Science
at the
University of Connecticut
2014
APPROVAL PAGE
Master of Dental Science Thesis
Presented By
Major Advisor:_____________________________________________
Dr. Gian Pietro Schincaglia, DDS, PhD
Associate Advisor:___________________________________________
Dr. Anna Dongari-Bagtzoglou, DDS, MS, PhD
Associate Advisor:___________________________________________
Dr. Patricia I. Diaz, DDS, MSc, PhD
Associate Advisor:___________________________________________
Dr. Thomas D. Taylor, DDS, MSD
ii
ACKNOWLEDGEMENTS
I wish to express my sincere gratitude to my major advisor, Dr. Schincaglia. I thank him
for his limitless dedication to teaching, leading, and mentoring me these past three years.
I will hold dear the lessons he has instilled throughout my entire career.
I want to thank all the members of my advisory committee, Dr. Dongari-Bagtzoglou, Dr.
Diaz, and Dr. Taylor. I thank them for their continued guidance throughout my time at
UConn and throughout the process of my Masters research.
I thank the entire faculty in the Division of Periodontology for their mentorship,
motivation, and support over the past three years.
My co-residents have become a second family to me. I thank them for their friendship
and support throughout my residency.
I am eternally grateful to my wife, Allison, who is the most amazing woman I have ever
known. Her love and support has made it possible for me to complete my educational
pursuits. I thank my amazing daughter, Toyba, for enriching every day of my life.
iii
5. Results .......................................................................................................................... 36
6. Discussion..................................................................................................................... 47
7. Conclusion ................................................................................................................... 52
8. References .................................................................................................................... 53
iv
1. Introduction
Over the past 20 years, implant therapy has been used as an alternative to tooth
replacement, to retain removable appliances, and to support fixed prostheses. Implant
therapy has become the gold standard in tooth replacement with over four and a half
million implants placed each year across the globe. While osseointegration of dental
implants occurs in more than 90% of the cases[1], once exposed into the oral cavity,
implants will be immediately colonized by bacterial plaque[2]. The consequent
interaction of the bacterial biofilm with the peri-implant soft tissue interface results in an
inflammatory reaction. The inflammatory state around the implant mucosa has been
termed mucositis. The periodontal literature has extensively characterized the tissue
response to the biofilm formed around the natural dentition. However, it is still unclear
whether the tissue response to bacterial biofilm around the natural dentition differs to that
around implants.
the immune response[10], the quality and quantity of the inflammatory infiltrate[8], and
the quality of the microbial flora[11]. This study utilizes a human model of experimental
peri-implant mucositis to evaluate the progression of the inflammatory process of the soft
tissue around implants and to compare that to the inflammatory process of gingiva around
teeth. This thesis seeks to clarify how the peri-implant mucosa and gingiva compare in
response to undisturbed plaque accumulation, using clinical indices of health and disease.
The 3rd National Health and Nutrition Examination Survey (NHANES III) conducted in
1988-1991 gives us some measure to determine the prevalence of gingivitis. However,
NHANES III did not directly attempt to quantify gingivitis. The data collected included
bleeding on probing, which we can then correlate to a moderate inflammation using the
index for assessment of gingivitis introduced by Le and Silness[15]. The periodontal
evaluation in NHANES III included a total of 7,447 dentate individuals 13 years and
older. The examinations were partial-mouth periodontal exams, performed on two
randomly selected quadrants, on 2 sites per tooth, with a maximum of 14 permanent teeth
eligible for assessment (third molars were excluded). Brown et al. [13] reported that
bleeding on probing was found in 62.9% for all persons aged 13 and older. However,
these data have to be considered with caution, since a partial-mouth periodontal
assessment has been shown to underestimate the prevalence of the disease[16].
Furthermore, the criteria to detect gingivitis include the visible signs of color change and
texture. Thus, a large percentage of individuals with gingivitis without bleeding on
probing were not included in the above statistic. A recent study by Li et al. [14] sought to
determine the prevalence of gingivitis in American adults using the Le and Silness
Gingival Index (GI) and found that 93.9% of individuals had a mean GI > 0.50. This
percentage decreased to 55.7% for individuals with a mean GI 1.0. These studies
suggest that the prevalence of gingivitis in the American population ranges between
55.7% and 93.9% depending on the definition used and method of calculation.
2.1.2 Etiology
The etiology of gingivitis was elegantly demonstrated in 1965 by Le [17] (see section
2.1.4). In this experiment, a direct cause and effect relationship was found between
plaque accumulation and gingival inflammation. Hence, this study offered proof that
bacterial biofilm causes gingivitis[17].
Under experimental gingivitis conditions, a significant shift was observed in the bacterial
composition from health to gingivitis[17]. The bacteria of healthy gingiva were
composed predominately of Gram-positive cocci and rods. When localized gingivitis
was first clinically detectable (GI = 1 on one tooth surface), Gram-negative cocci and
rods were found in increasing percentages in the microflora. There was a shift again as
generalized gingivitis was clinically identified (GI = 1 for the individual). Here, Grampositive and negative filaments, fusobacteria, spirilla, and spirochetes were more readily
measurable. In the study by Theilade et al. [18] a correlation between gingival status and
plaque composition was established. Theilades work details the growing complexity of
the microflora composition as plaque accumulates[18, 19]. The authors results indicate
that gingivitis is not due to an infection by a single pathogen, but rather a plaque ecology
that is unfavorable to the host[20].
2.1.3 Histopathology
The histological features of healthy gingiva include a keratinized oral epithelium
continuous with a junctional epithelium (JE) that is attached to the tooth surface by
hemidesmosomes. Polymorphonuclear leukocytes (PMN), while small in overall number,
are present in the healthy sulcus as a normal reaction to the biofilm present around teeth
even in a meticulously cleaned mouth[21]. If sufficient plaque is allowed to accumulate,
gingivitis will ensue. Page and Schroeder described the pathogenesis of gingivitis and
periodontitis as a continuum of increasing inflammatory response by the host to microbial
plaque[21].
The first stage of gingivitis is the initial lesion. It is formed within 2-4 days of the
initiation of plaque accumulation. The characteristics of the initial lesion are vasculitis
adjacent to the JE, increased flow of gingival crevicular fluid (GCF), and increased
migration of PMNs into the JE and sulcus. Additionally, an altered morphology of
epithelial cells in the coronal part of the JE and loss of perivascular collagen can be seen.
Neutrophils adhere to vessel walls and migrate to the connective tissue, JE, and sulcus.
At this stage bacterial invasion is not evident and the inflammatory response is thought to
be related to plaque-derived chemotactic substances[21].
The second stage of gingivitis is called the early lesion. It is mostly an accentuation of
the initial lesion and occurs at days 4 through 7. There is an accumulation of lymphoid
cells (approximately 75% of the total infiltrate) adjacent to the JE, further loss of collagen
fiber network supporting the marginal gingiva and the beginning of proliferation of the
basal cells of the JE. The primary distinguishing feature here is the migration of
lymphoid cells (predominately T cells with very few B cells) into the lesion[21].
The third and final stage of gingivitis is the established lesion. It is formed at least 7 days
after the initiation of plaque accumulation. The histo-pathological hallmark of this
condition is the predominance of plasma cells. Immunoglobulins are seen
extravascularly in the connective tissues and JE. There is a continuing loss of connective
tissue matrix elements and proliferation, apical migration, and lateral extension of the JE
leading to the formation of a pocket epithelium[21]. This established lesion can remain
stable for months or years without progressing. The established lesion is still reversible.
Further persistence of the established lesion may eventually progress to the advanced
lesion, which is defined as irreversible and accompanied by attachment loss. Once
attachment loss occurs, the condition is defined as periodontitis.
The experimental gingivitis model has been used ubiquitously over the past 50 years to
measure a variety of outcome variables. One of these outcome variables is whether age
plays a role in susceptibility to gingivitis. In an experiment by Holm-Pedersen et al. [22],
gingivitis developed more rapidly and more severely in elderly than in young persons.
The authors suggest that with age comes an altered host response to the bacterial plaque.
The effect of age on the host response to plaque accumulation was confirmed by Matsson
[23], who compared pre-school aged children with young adults. The author found that
the younger group had a statistically significantly lower amount of gingival exudate and
gingival bleeding than the older group. Winkel et al. [24] tested the effect of age on the
rate of development of gingivitis in persons not susceptible to periodontal destruction.
The findings of this study were that in individuals not susceptible to periodontal
destruction, age was not a factor in the development of experimentally induced gingivitis.
Winkel et al. and Matsson use different classifications of age, making direct comparisons
of their studies difficult. Thus, there may be a critical age difference required to detect a
true difference in gingivitis susceptibility to plaque accumulation.
6
The experimental gingivitis model has also been used to elucidate the histopathologic and
immunologic features of gingival inflammatory changes. Payne et al. [21, 25] used the
experimental gingivitis model to describe the pathologic alterations in gingival tissues in
early plaque accumulation. They found that within the first 8 days of plaque
accumulation, a lesion resembling a delayed hypersensitivity developed. Seymour et al.
[26] took biopsies at four time points through a 21-day oral hygiene abstention study and
found that the nature of the composition of the infiltrate did not change significantly, as
the degree of inflammation increased.
Another use of the experimental gingivitis model was to evaluate techniques and
procedures to reverse gingivitis or to mitigate its onset. Bosman and Powell [27]
demonstrated that a once daily rinse with 0.2% chlorhexidine (CHX) mouthrinse regained
gingival health in 4 days versus 10 days when only brushing was performed. Gusberti et
al. [28] investigated the ability of two therapeutic mouth rinses, 0.12% CHX and 1%
hydrogen peroxide (H2O2), to mitigate the development of gingivitis when all other oral
hygiene measures were removed. They found 0.12% CHX to significantly reduce
gingivitis incidence, plaque, bleeding on probing, and specific facultative and obligate
anaerobes compared to a placebo. Conversely, 1% H2O2 did not have a significant
difference over the placebo[28]. Ultimately, these types of studies aid in distinguishing
the therapeutic value of different chemotherapeutic agents in treating gingivitis.
The early experimental gingivitis studies hinted at the possibility of differences in subject
susceptibility to plaque accumulation. This response variability was interpreted as related
to differences in plaque accumulation rates or differences in microbial composition[17,
18]. More recently, Trombelli et al. [29] in an experimental gingivitis trial were able to
identify two subgroups of subjects that differed in their response to similar plaque
exposure. These were labeled the High Responder (HR) and Low Responder (LR).
The identification of these two subpopulations was based on their amount of GCF
volume. This finding supports the hypothesis of the existence of host-related factors
implicated in the individual variability of the inflammatory response to plaque. The
implication of susceptibility to gingivitis due to a subjects genetic make-up or host
response versus a purely microbial reaction has changed how the disease process is
studied. More efforts are looking into the effect of genetic variation and other factors that
could further elucidate the pathogenesis of gingivitis.
Prevalence values were taken from Table 8, where BOP was detected at implant level or
patient level at implants where no bone loss was identified between 1-year radiographic
control and final examination. If prevalence was determined by presence of Bleeding on
Probing and Pocket Probing Depth >/= 4mm irrespective of bone levels then prevalence
would be 76.6% at the patient level and 48.1% at the implant level (Table 6).
1
peri-implant mucositis on implants with no additional bone loss after baseline (n=285
implants) was 90.9% at implant level analysis.
Roos-Jansker et
al. [4, 32]
Ferreira et al.
[33]
Karbach et al.
[34]
Koldsland et al.
[35, 36]
Maximo et al.
[37]
A recent systematic review and meta-analysis by Atieh et al. [41] found 63.4% of
subjects and 30.7% of implant sites had peri-implant mucositis. As demonstrated, the
prevalence of peri-implant mucositis cited in the literature varies dramatically. This
variability could be explained by the difference in the methods to diagnose this condition.
2.2.2 Etiology
Peri-implant mucositis is caused by bacterial plaque. To establish the relationship
between plaque formation and inflammation around implants, Pontoriero et al. [7]
conducted a three-week experimental gingivitis peri-implant mucositis study in partially
edentulous subjects restored with dental implants. Phase contrast microscopy was used
to evaluate submucosal and subgingival plaque samples. Coccoid cells, motile rods, and
spirochetes were found in similar proportions at baseline and at the end of the three-week
experimental period on both the implants and teeth. The period of oral hygiene
abstention demonstrated the same cause and effect relationship between plaque
accumulation and inflammation for implants and mucositis as for teeth and gingivitis.
The results were similar to those found decades earlier by Le et al. [17] and Theilade et
al.[18]. A more recent study by Salvi et al. [10] also confirmed the cause and effect
relationship between plaque formation and inflammation around the soft tissues of
implants. In this study the peri-implant mucosa appeared to have a stronger response to
plaque compared to the gingiva. Nonetheless, a well-established causal relationship
between plaque accumulation and peri-implant mucosal inflammation has been
consistently validated.
10
Oral bacteria colonize on the implant as early as 30 minutes to two weeks following
implant placement. Furthermore, the microbial flora at implants presents a composition
almost identical to the adjacent teeth[42, 43]. This may suggest that in partially dentate
patients, adjacent pockets harboring periodontal pathogens can act as a reservoir for
colonization and establishment of a microflora that may not be supportive of mucosal
health[44].
Colonization of the implant occurs similarly to that of a tooth. When the implant is
exposed into the oral environment, a salivary pellicle forms on the surface. The pellicle
forming on a titanium surface is similar to that forming on enamel, including salivary
alpha-amylase and proline-rich proteins[45]. The pristine pockets of a newly inserted
titanium implant abutment are not only colonized rapidly, but with a complex subgingival
microflora that mirrors the subgingival plaque of shallow gingival crevices[2].
There are two differences between bacterial colonization on teeth and implants that have
been identified thus far. First, the numbers of the red and orange complex bacteria, while
present, appear to be lower on implant sites than for tooth sites[2]. Quirynen et al. [2]
explained this by suggesting that some taxa are able to colonize a clean site on their own,
while other taxa require the establishment of appropriate conditions by pioneer species.
In addition, Buchmann et al. [46] found that implants can be colonized by a complex
microflora that included Porphyromonas gingivalis and Aggregatibacter
actinomycetemcomitans. However, this microbial colonization was compatible with periimplant mucosal health. The second difference in bacterial colonization of teeth and
implants is the specific bacterial profile in the presence of periodontitis and peri-
11
12
(also attaching via hemidesmosomes to a titanium oxide implant surface [50]) that is
between 1 to 1.5 mm long. While the principal fibers of teeth invest in the root
cementum and fan out in lateral, coronal and apical directions from the root, the implant
equivalent fibers invest in the bone crest and run parallel to the implant[51]. Differences
in the vascular supply are also evident between the gingiva and the peri-implant mucosa.
The vascular supply for the gingiva originates from the periosteum and the periodontal
ligament. The vascular supply for the peri-implant mucosa comes solely from the
periosteum[52]. Leukocytes are present in small numbers in a normal reaction to the
biofilm present around implants, which is similar to that found in healthy gingiva[9]. The
connective tissue around implants was found by Berglundh et al. [9] to have greater
collagen and lower density of fibroblasts than the corresponding gingival connective
tissue. Despite these noted differences, the peri-implant mucosa and gingiva are
remarkably similar.
As outlined previously for teeth, if sufficient plaque is allowed to accumulate around the
implants, mucositis will ensue. Animal[5, 9, 53] and human[7, 8] studies have shown a
comparable reaction to an experimental biofilm accumulation in gingiva and soft tissues
adjacent to dental implants, when examined both clinically and histologically. This is
described in more detail in section 2.2.4.
13
Using a similar experimental model, Berglundh et al. [9] evaluated the changes of
gingiva and peri-implant mucosa after 21 days of plaque accumulation. The peri-implant
mucosa and gingiva had similar reactions to early plaque formation. Both had an
increase in leukocyte transmigration, similar apical extensions of the ICT, and similar
14
volumes of inflammatory cells in the connective tissue lesions. The results of the study
point out that teeth and titanium abutments acquire similar plaque quantity after 21 days
of accumulation, and that both gingiva and peri-implant mucosa respond similarly to the
accumulated plaque[9].
The beagle dog model was used to explore the nature of lesions resulting from
subgingival and submucosal ligature placement for six weeks[5]. Clinical evaluation
demonstrated that signs of tissue destruction were more pronounced at implants than
teeth. Biopsies revealed that the implants had a larger soft tissue lesion than teeth. The
inflammatory infiltrate on teeth was contained in the connective tissue whereas lesions on
the implant site extended into the bone marrow [5]. This study suggests that long lasting
inflammation in the peri-implant mucosa can progress faster into the supporting bone
when compared to teeth.
In summary, the series of experiments utilizing the dog model indicate that peri-implant
mucosa and gingiva respond similarly under conditions of plaque accumulation from a
clinical perspective. However, structural differences between the peri-implant mucosa
and gingiva may render the peri-implant mucosa less effective in preventing the
progression of an inflammatory infiltrate.
15
Similarly, Zitzmann et al. [8] investigated the soft tissue reactions to de novo plaque
accumulation over a three-week period in humans. In this study, 12 subjects, each with
two implant sites and two control teeth sites, had two biopsies taken at baseline and after
three weeks of plaque formation. From these biopsies, histological, histometric, and
immunohistochemical evaluations were performed. The clinical parameters evaluated
(PlI and mGI) were not different between tooth and implant sites. On histological
observations both gingiva and peri-implant mucosa had large cell infiltrates present in the
connective tissue lateral to the junctional epithelium. Comparisons were made with
histomorphometry and immunohistochemistry. No differences in the infiltrate between
the gingiva and the peri-implant mucosa were observed at any time point. However, the
authors point out that mean cell densities in the infiltrate of the gingiva were consistently
higher than in the peri-implant mucosa[8]. Zitzmann and Pontoriero both support the
concept that both gingiva and peri-implant mucosa have a similar inflammatory response
to plaque accumulation.
Salvi et al. [10] used an experimental gingivitis model to clarify the pathogenesis of
gingival/mucosal inflammation with regard to specific clinical, microbiological, and hostderived factors. As stated earlier, they found a comparable cause and effect relationship
regarding bacterial challenge and host response around both natural teeth and implants.
A stronger inflammatory response to plaque accumulation was seen in the mucosa
surrounding implants versus natural teeth (i.e. statistically significant increase in GI,
16
Pocket Probing Depth [PPD], and MMP-8). No difference between the subgingival and
the submucosal microbiota was observed. Additionally, the authors examined the
response to reinstitution of oral hygiene measures on the resolution of inflammation.
Reversibility was seen at the biomarker level for both mucositis and gingivitis. The
authors found that 3 weeks of resumed plaque control appears insufficient to bring
gingival and peri-implant mucosa back to baseline levels[10].
Two methods have been utilized for quantitating GCF volume: the ninhydrin area
method (NAM) and the Periotron. The NAM relies on staining serum alpha-amino
proteins absorbed on a filter paper with ninhydrin and quantifying the resulting stained
area. The Periotron measures electrical capacitance of the fluid absorbed on the filter
paper strip, which changes in dielectric insulating properties as the quantity of fluid
absorbed changes[57]. There have been several versions of the Periotron: the Periotron
17
600, Periotron 6000, and the most recent Periotron 8000. The Periotron is most
appropriate to evaluate longitudinal changes of GCF in a particular individual, but not
between individuals[62].
BOP has been also evaluated around dental implants. In an experimental peri-implant
mucositis and experimental peri-implantitis study, Lang et al. [64] demonstrated that
healthy peri-implant sites had zero BOP, sites with experimental peri-implant mucositis
had a mean BOP of 66.7%, and sites with experimental peri-implantitis had a mean BOP
of 90.9%. Thus, destruction of peri-implant tissues was found to be associated with BOP.
To answer the question of how reliable BOP is as a predictor of future attachment loss
around implants, Jepsen et al. [65] evaluated 25 patients with 54 implants (mean loading
time of 41 months). A minimum threshold of 1mm of attachment loss was selected to be
positive for site breakdown. BOP was found to have a high negative predictive value for
peri-implant attachment loss (82% at the implant level and 97% at the site level).
Similarly to teeth, the absence of BOP around implants can serve as an indicator for
18
stable peri-implant conditions[65]. When the threshold for disease progression is defined
as an annual increase in PPD 0.5mm or 2.5mm in 5 years, BOP at 50% of sites had a
specificity of 100% (compared to 73% for teeth) and a positive predictive value of 100%
(compared to 40% for teeth)[66].
Thus, these results seem to indicate that BOP may be a predictor for peri-implant bone
loss. However, gingival and peri-implant mucosa are structurally different and lighter
pressure during probing around implants has been recommended. Gerber et al. [67]
showed the mean BOP percentage at implants (13.7%) and teeth (6.6%) increase as the
probing pressure increased from 0.15N to 0.25N. The authors conclude that a probing
force of 0.15N should be applied around implants to avoid false positive BOP[67]. While
BOP as a predictor of disease activity is limited, it may be the most useful clinical
predictor we have during implant maintenance programs.
There are multiple methods to measure gingival and peri-implant mucosal bleeding due
to inflammation. The Angulated Bleeding Index (AngBI) [68], the method by which the
periodontal probe is placed to the level of the marginal gingiva at a 60-degree angle, is
one example. In an experimental gingivitis model, this method has been shown to be a
sensitive indicator of early changes in gingival inflammation [68]. A modification of this
19
technique was recently described by Trombelli et al. [29] and is known as the Angulated
Bleeding Score (AngBS). While this method has been used in several experimental
gingivitis studies[29, 68], it has yet to be employed in an experimental peri-implant
mucositis model.
0 = Negative. There is neither overt inflammation in the investing tissues nor loss
of function due to destruction of supporting tissue.
1 = Mild Gingivitis. There is an overt area of inflammation in the free gingiva,
but this area does not circumscribe the tooth.
2 = Gingivitis. Inflammation completely circumscribes the tooth, but there is no
apparent break in the epithelial attachment.
6 = Gingivitis with Pocket Formation. The epithelial attachment has been broken
and there is a pocket (not simply due to swelling of the free gingiva). There
is no interference with normal masticatory function, the tooth is firm in its
socket, and has not drifted.
8 = Advanced Destruction with Loss of Masticatory Function. The tooth may be
loose; may have drifted; may sound dull on percussion with a metallic
instrument; and may be depressible in its socket.
Essentially a mean score below 2 would be considered in a range of health to gingivitis,
while a score above 2 would be considered periodontitis.
20
The most widely used and accepted classification system was developed by Le and
Silness[15]. The criteria for the GI System proposed was as follows:
0 = Absence of inflammation.
1 = Mild Inflammation. Slight change in color and little change in texture.
2 = Moderate Inflammation. Moderate glazing, redness, edema, and hypertrophy.
Bleeding on pressure.
3 = Severe Inflammation. Marked redness and hypertrophy. Tendency to
spontaneous bleeding. Ulceration.
In an effort to define parameters more appropriate for peri-implant mucosa, Mombelli et
al. [70] applied a modified Sulcus Bleeding Index (mBI), which is often referred to as
modified Gingival Index (mGI), to describe the level of inflammation. In other words,
the mBI replaces GI for implants. The criteria for the mBI system proposed was as
follows:
0 = No plaque.
1 = A film of plaque adhering to the free gingival margin and adjacent area of the
tooth. The plaque may be seen in situ only after application of disclosing
solution or by using the probe on the tooth surface.
2 = Moderate accumulation of soft deposits within the gingival pocket, or on the
tooth and gingival margin which can be seen with the naked eye.
3 = Abundance of soft matter within the gingival pocket and/or on the tooth and
gingival margin plaque.
Mombelli et al. [70] also supplied a modification to the Plaque Index (mPlI) in order to
make it more fitting for peri-implant mucosa. The criteria for the mPlI system proposed
was as follows:
0 = No detection of plaque.
1 = Plaque only recognized by running a probe across the smooth marginal
surface of the implant. Implants covered by titanium spray in this area
always score 1.
22
23
24
analysis by Schou et al. [75], the authors found a significant increase in the number of
patients with peri-implantitis after a 10-year follow-up, when their tooth loss was due to
periodontal disease, risk ratio (RR) 9 (95% CI 3.94-20.57). Marginal bone loss after 5
years around implants in patients whose tooth loss was due to periodontitis was
significantly increased compared to patients whose tooth loss was due to other reasons.
Ferreira et al. [33] reported an OR of 3.1 (95% CI 1.1-3.5) for patients with a history of
periodontitis to develop peri-implantitis. This relationship was not found to be
statistically significant for peri-implant mucositis. In their five-year retrospective study
of patients diagnosed with peri-implant mucositis at baseline, Costa et al. [55] found the
OR for these patients to develop peri-implantitis was 9.2 (95% CI 1.53-55.32) if they
currently had periodontitis. The OR increased to 11.43 (95% CI 1.11-117.33) if these
patients with periodontitis were not in a routine maintenance program.
Simonis et al. [76] evaluated the long-term results of implants in 55 patients with 131
Straumann tissue level implants after 10-16 years. The long-term survival rate was
82.94%, with frequent biological (mucositis and implantitis) and technical complications.
Patients with a history of periodontitis showed a prevalence of peri-implantitis of 37.93%,
while in subjects without a history of periodontitis, peri-implantitis was present in
10.53% of the cases. This translated to an OR of 5.1 (95% CI 1.92-14.06). The literature
consistently found that subjects with history of periodontitis were much more likely to
develop peri-implant disease than persons without periodontitis.
25
2.4.3 Smoking
Smoking is a risk factor extensively evaluated in the implant literature. Several studies
have demonstrated that smokers have significant increased risk for peri-implant
mucositis[34, 77, 78]. Karback et al. [34] reported smoking to be associated with perimplant mucositis with an OR of 3 (95% CI 1.141-7.916). In a prospective study of 24
implants in 10 smokers and 18 implants in 4 nonsmokers of at least one-year follow-up,
Ataoglu et al. [79] found that smokers had a significant increase in the inflammationrelated clinical parameters PPD, mPI, mGI, and GCF flow rate.
A retrospective study by Hass et al. [80] evaluated 1366 implants in 421 patients (366
implants in 107 smokers and 1,000 implants in 314 nonsmokers) with an average recall
time of 22 months. Smokers were found to have increased bleeding index, PPD, degree
of peri-implant mucosal inflammation, and radiographic bone loss. These differences
were significant in the maxilla compared to mandible in smokers and in the maxilla
between smokers and nonsmokers. These findings may indicate that implants in the
maxilla are more susceptible to the effects of smoking than in the mandible.
26
review by Strietzel et al. [81] found implant failure in smokers had an implant related OR
of 2.25 (95% CI 1.96-2.59), and a patient related OR of 2.64 (95% CI 2.26-5.77). Thus
indicating that smokers had a significantly increased risk of biological complications
around implants.
2.4.4 Diabetes
The risks of implant failure and biologic complications associated with diabetes are more
controversial in the literature. This may be due to a significant heterogeneity between
studies that investigated the effects of diabetes on implant health. While some studies
investigate the implant outcomes for patients with poor diabetic control[33], others
evaluated whether there are differences in implant success in patients with good diabetic
27
control[84]. Lastly, some studies do not specifically note the level of diabetic control of
the patients[85-87]. With regard to patients with good diabetic control, Abdulwass and
Dhanrajani [84] reported no increased risk of implant failure in patients that had HgA1c
levels 7% and are in a supportive periodontal therapy program, when compared to nondiabetics. Conversely, Ferreira et al. [33] showed that patients with poor diabetic control
have an increased risk of peri-implant mucositis and peri-implantitis.
In studies where the level of diabetic control is not reported, the conclusions are
controversial[85-87]. Using the results of the two studies described above as examples,
well-controlled diabetes does not appear to cause increased risk of biologic complications
around implants, while uncontrolled diabetes is a risk factor for peri-implant disease. By
not stratifying this variable, the true effect of diabetes on implants cannot be fully
elucidated.
3.2. Hypothesis
There is a different clinical response to plaque accumulation in peri-implant mucosa
when compared to gingiva.
28
3.3 Objectives
To assess changes in Plaque Index, Gingival Index, Angulated Bleeding Score, and
Gingival Crevicular Fluid volume between peri-implant mucosa (test) and gingiva
(control) over the two experimental phases (stimulation and resolution of inflammation).
29
30
31
Screening visit: The experimental procedure was explained and a written informed
consent was presented for patient acceptance. Inclusion and exclusion criteria were
verified with a medical questionnaire, clinical oral exam, and dental radiographs. The
radiographs, taken only if current (within 1 year) films were not readily available for
review, were used to determine the presence or absence of bone loss around implants and
teeth. The qualified patients were then provided with 56 tablets of 250 mg vitamin C
supplement to be taken daily during the pre-trial and trial period. The purpose of the
32
vitamin C was to prevent any possibility of a deficiency that could affect the patients
susceptibility to inflammation.
Visit 1 (day -14): General OHI was provided to each patient. A toothbrush, floss, and
fluoridated toothpaste were provided to each patient (Colgate-Palmolive Company, New
York, NY, USA). Peripheral blood (10 mL) was collected in EDTA tubes and
immediately stored at -80C until further analysis (results to be reported elsewhere).
Visit 2 (day -7): All patients received a review of OHI, stent trial, and a second round of
scaling and polishing if needed. Experimental procedures were reviewed.
Visit 3 (day 0): Healthy gingiva (GI = 0) and excellent plaque management (PlI = 0) was
verified. If subjects did not satisfy this prerequisite, they were provided another round of
scaling and polishing and oral hygiene instructions and asked to return in one week.
Visit 6 (day +21): Stents were collected and patients were instructed to reinstitute oral
hygiene. A new toothbrush and fluoridated toothpaste were provided to each patient
(Colgate-Palmolive Company, New York, NY, USA).
Visit 7 (day +42): Scaling and polishing and a review of oral hygiene instructions
followed collection of clinical parameters.
33
Clinical Parameters
Clinical parameters were collected as listed below from the test and control sites selected
for each patient.
GCF was collected as described by Trombelli et al. [29] and measured using the
Periotron 8,000 (OraFlow Inc., Plainview, NY, USA). The specific site was
isolated using a cotton roll to avoid salivary contamination and gently air dried in
an apico-coronal direction. Supragingival plaque was not disturbed. A sterile
paper strip (Periopaper, OraFlow Inc,) was placed into the sulcus until mild
resistance was sensed. The paper strip was held in place for five seconds and
immediately measured chair-side by the calibrated Periotron 8,000.
PlI according to Silness and Le [71] and applied by the method described by
Furuichi et al. [88]. Plaque was first identified as either visible or not visible. If
visible, the plaque was categorized as either a score of 2 or 3. If not visible, a PlI
score of 0 or 1 was determined by staining the tooth or implant with erythrosine
solution (Red Cote, Butler, Chicago, IL, USA) using a syringe applicator.
34
Stent Fabrication
Cast models were prepared from the alginate impressions taken on visit 1. Wax was
placed (2 mm thick) from coronal to the height of contour to 4-5 mm apical to the
gingival/mucosal margin on both the buccal and lingual/palatal surfaces of the test and
control sites. This was done to allow the plaque to accumulate undisturbed. A duplicate
of this altered model was made. Stents were fabricated using 1.5 mm thermo-forming
material (Henry Schein Inc., Melville, NY, USA) and a vacuum forming machine (Henry
Schein Inc., Melville, NY, USA). The stents were trimmed and adjusted as needed in the
patient on visit 2. Patients were instructed to insert the stent prior to brushing during the
experimental phase. This allowed the remainder of the dentition to be brushed without
removal of plaque from the experimental sites.
35
mucositis studies in humans [7, 8, 10]. Considering 20% attrition, a total of 18 subjects
were enrolled.
The null hypothesis was that no significant clinical differences assessed during the
experimental three-week period of absent oral hygiene would exist between the periimplant mucosa and the gingiva. The Kolmogorov-Smirnov goodness fit test was
computed to test the normal distribution for each parameter. All the variables were found
not normally distributed. Therefore, inter-individual comparisons were performed using
the Mann-Whitney Test. The Wilcoxon Signed Rank Test was used to analyze intraindividual comparisons. The Spearman Rank Correlation Test determined correlations
between the four clinical parameters. Significance for all variables was set at 95%
(=0.05).
5. Results
Demographics and protocol deviations
Eighteen patients were enrolled. For the purpose of this report we are presenting the
results relative to 14 completed subjects. One participant has not yet completed the trial.
During the study, 3 patients exited the experiment. Patient flow is reported in Fig. 2.
One patient was excluded from the study due to the development of an acute episode of
swelling of the mucosal margin of a test site with drainage observed at the d.14 visit. The
site was immediately treated with debridement of the area, thorough irrigation and one
week of 0.12% chlorhexidine mouth rinse. The condition completely resolved at one
week. The other two participants were excluded from the study due to a change in
medical condition unrelated to the study proceedings.
36
Plaque Accumulation
The clinical data of the median and range of the PlI are presented in Table 4. Both
control and test groups showed a statistically significant decrease in PlI from the nave
state (d.-14) to day 0 (d.0) (p=0.002 and p=0.008, Wilcoxon Signed Rank Test).
Compared to d.0, there was a statistically significant increase in PlI at each time point of
37
plaque accumulation for the control and test groups over the experimental phase (Table 5,
Fig. 3a and b). There was statistically significantly greater plaque on d.42 than on d.0 in
both groups (p=0.004 control and p=0.008 test, Wilcoxon Signed Rank Test). In neither
group was there a statistically significant difference between the d.-14 and d.42. The
control group had a significantly greater PlI than the test groups on d.14 and d.21
(p=0.036 and p=0.048, Mann-Whitney Test) (Table 5 and Fig. 4).
Table 4. Clinical data for PlI, GI, AngBS and GCF for control and test sites.
38
Table 5. Within group and between group (shaded) comparisons among median PlI
measured at different time points in control and test groups.
*Not significant. Significance level: < 0.05. Mann-Whitney Test was used to determine
significance between groups (shaded) and Wilcoxon Signed Rank Test was used for
within group analysis.
39
b
Fig. 3. Box-Whisker plot for plaque index over pre-experimental, experimental, and
resolution phases in control (a) and test (b) groups (n =14).
Fig. 4. Box-Whisker plot for plaque index over pre-experimental, experimental, and
resolution phases comparing control and test groups (n =14). *Significance level: < 0.05,
Mann-Whitney Test used for between group analysis.
Gingival Inflammation
The clinical data of the median and range of the GI are presented in Table 4. Both groups
showed a statistically significant decrease in GI from the d.-14 to d.0 (p=0.016, Wilcoxon
40
Signed Rank Test). Compared to d.0, there was a statistically significant increase in GI at
each time point for the control and test groups over the experimental phase (Table 6 Fig.
5a and b). The GI was statistically significantly greater on d.42 than on d.0 in both
groups (p=0.016 control and p=0.031 test, Wilcoxon Signed Rank Test). In neither group
was there a statistically significant difference between the d.-14 and d. 42. Significant
differences were not observed between the control and test groups at any time point
(Table 6 and Fig. 6).
Table 6. Within group and between group (shaded) comparisons among mean GI
measured at different time points in control and test groups.
*Not significant. Significance level: < 0.05. Mann-Whitney Test was used to determine
significance between groups (shaded) and Wilcoxon Signed Rank Test was used for
within group analysis.
41
b
Fig. 5. Box-Whisker plot for gingival index over pre-experimental, experimental, and
resolution phases in control (a) and test (b) groups (n =14).
42
Fig. 6. Box-Whisker plot for gingival index over pre-experimental, experimental, and
resolution phases comparing control and test groups (n =14).
The clinical data of the median and range of the AngBS are presented in Table 4. The
control group demonstrated a statistically significant increase in AngBS from d.0 to d.7,
d.14, and d.21 (p=0.016, 0.031, and 0.008, Wilcoxon Signed Rank Test). The test group
demonstrated a statistically significant increase from d.0 to d.14 and d.21 (p=0.016 and
0.018,Wilcoxon Signed Rank Test). There were no other statistically significant
differences in AngBS within groups over time (Fig. 7a and b). The test group had a
significantly greater AngBS than the control on d.42 (p=0.043, Mann-Whitney Test) (Fig.
8).
43
b
Fig. 7. Box-Whisker plot for angulated bleeding score over pre-experimental,
experimental, and resolution phases in control (a) and test (b) groups (n =14).
*Significance level: < 0.05, Wilcoxon Signed Rank Test used for within group analysis.
44
The clinical data of the median and range of the GCF are presented in Table 4. The
control group demonstrated a statistically significant increase in GCF between the d.-14
and d.21 (p=0.023, Wilcoxon Signed Rank Test). Significant differences were also seen
between multiple time intervals (Fig. 9a). The test group demonstrated a statistically
significant increase from d-14 to d.7 and d.14 (p=0.016 and 0.044, Wilcoxon Signed
Rank Test). Additionally, statistical differences were observed from d.0 to each of the
other time intervals in the experimental phase (Fig. 9b). The test group had a
significantly greater GCF than the control on d.7 (p=0.038, Mann-Whitney Test) (Fig. 10).
45
b
Whisker plot for gingival crevicular fluid volume over pre-experimental,
experimental,
Fig. 9. Box-Whisker
experimental, and resolution phases in control (a) and test (b) groups ((n =14).
*Significance level: < 0.05, Wilcoxon Signed Rank Test used for within group analysis.
analysis
46
Fig. 10. Box-Whisker plot for gingival crevicular fluid over pre-experimental,
experimental, and resolution phases comparing control and test groups (n =14).
*Significance level: < 0.05, Mann-Whitney Test used for between group analysis.
Correlations between PlI, GI, AngBS, and GCF volume were analyzed. A correlation
between GI and AngBS was found at d.21 for the test group (p=0.026, rs=0.609,
Spearman Rank Correlation Test). Also in the test group, a correlation between GI and
PlI was seen at d.7 (p=0.02, rs=0.625, Spearman Rank Correlation Test). No other
correlations were observed between any of the clinical parameters at any time point in
either the test or the control groups.
6. Discussion
The primary objective of this study was to assess changes in the PlI, GI, AngBS, and
GCF volume between gingiva and peri-implant mucosa over a 21 day period of plaque
accumulation (experimental phase), and following a 21 day period of resumption of oral
hygiene measures (resolution phase). The control and test groups demonstrated a
significant effect of time on PlI score. Statistically significant increases in PlI over the
47
experimental phase, and statistically significant decreases in PlI over the resolution phase
were observed. This finding demonstrates that the study methodology to induce plaque
accumulation was effective. When comparing the two groups, the PlI was significantly
higher in the control group at d.14 and d.21 when compared to the test. This result was
expected. Ceramo-metal restoration have showed lower plaque accumulation when
compared with unrestored teeth[89]. Natural teeth accumulate more plaque than crowned
teeth under the same conditions due to their differences in surface roughness. The lower
PlI in the test sites was observed despite the presence of full coverage crowns in 50% of
the control sites. Additionally, the median PlI scores on teeth and implants in our study
were consistent with PlI scores reported in the other experimental peri-implant mucositis
studies in humans[7, 8, 10].
This study sought to investigate differences between gingiva and peri-implant mucosa in
their responses to plaque accumulation. The GI, AngBS, and GCF volume are indirect
measure of the peri-implant mucosal/gingival inflammatory response[90].
The control and test groups demonstrated a significant effect of time on GI score.
Statistically significant increases in GI over the experimental phase, and statistically
significant decreases in GI over the resolution phase were observed in both control and
test groups. However, there were no statistically significant differences seen in GI scores
between gingiva and peri-implant mucosa at any time point during the pre-experimental
phase, the experimental phase, or the resolution phase, respectively. One of the
limitations of this study is that for enrollment convenience we did not exclude subjects
with full crown coverage in the control sites. This may raise concern in relation to a
48
possible effect of the crown margin on the response of the gingival tissue. A
retrospective study by Carnevale et al. [91] assessed the PlI and GI of 510 crowned teeth
and 510 natural teeth in patients under a strict maintenance program. The results
demonstrated that there were no differences in PlI and GI between teeth with and without
crowns. This held true regardless of where the crown margin was placed (supragingival,
at the gingival margin, or subgingival).
The lack of significant difference in GI between test and control observed in our study
differed from the results reported by Salvi et al. [10]. In that study, statistically
significant differences were observed between the gingiva and peri-implant mucosa
during the experimental and resolution phases. The peri-implant mucosa had greater GI
than the gingiva. One reason for this difference could be related to the methodology used
to measure GI. The Salvi et al. [10] study utilized the mGI described by Mombelli et al.
[70], while in our investigation we utilized the GI method according to Le and Silness
[15] modified by Trombelli et al.[29]. We used this method because one of the priorities
in this study was to leave the plaque undisturbed throughout the whole experimental
phase. As detailed in section 2.3.3, by removing the bleeding component, the ability to
distinguish between a GI score of 1 or 2 is limited. Hence, the lack of significant
difference in GI should take into consideration that this methodology may have resulted
in a less sensitive assessment of GI.
The GI in both groups on d.42 was significantly greater than the GI on d.0 (baseline).
This result, also observed by others[10], suggests that the resolution of an established
inflammatory state may require a longer period of plaque control to return to baseline.
49
However, the GI in both groups on d.42 was not statistically different than the GI on d-14
(nave condition).
50
The GCF volume as a clinical parameter of gingival inflammation has been demonstrated
to have the highest correlation with PlI[29]. The GCF volume increased throughout the
experimental phase and decreased after the resolution phase. On d.7, the test group had a
significantly greater GCF volume compared to the control group. This could suggest that
the peri-implant mucosa responds more immediately to experimental plaque
accumulation than gingiva. The traditional method for collecting GCF from the gingival
or peri-implant mucosal sulcus involves the removal of supragingival plaque; after which
a sterile paper strip is placed into the sulcus for 5 to 30 seconds[60, 92]. Supragingival
plaque has been shown to significantly increase the GCF measurements in subjects with
healthy gingiva[60]. However, removing the supragingival plaque would not be
compatible with our study design of undisturbed plaque accumulation. Thus, the
potential influence of plaque on our GCF measurements cannot be ruled out. This is the
first study to our knowledge that directly compares the GCF in the peri-implant mucosal
sulcus to that of the gingival sulcus using this methodology. Further investigations are
needed to confirm these data.
One of the objectives in our study was to describe the clinical parameters between the
control and test groups in the nave condition (d.-14). For each clinical parameter
evaluated at this time point, no differences were observed between the two groups. This
observation was never reported in previous investigations. Previous experimental periimplant mucositis/gingivitis studies in humans all consider baseline to be after
51
professional debridement with oral hygiene instruction and motivation[7, 8, 10]. Their
aim was to have baseline PlI and/or GI scores close or equal to zero prior to commencing
the experimental phase. In our study, by evaluating participants at their first visit
comparisons could be made to each subsequent step in the experimental and resolution
phases. We found that while PlI and GI scores remained significantly elevated in both
control and test groups when comparing d.0 to d.42, there were no differences when
comparing d.-14 to d.42.
The lack of significant difference for the clinical parameters analyzed in our study must
be taken with caution due to the small sample size used. However, these data will be
utilized to determine the sample size in future studies.
7. Conclusion
In conclusion, the results from this study support a cause and effect relationship between
plaque accumulation and clinical parameters of gingival and peri-implant mucosal
inflammation. These results are in agreement with those found in several experimental
peri-implant mucositis studies in humans[7, 8, 10]. We showed that 21 days of
undisturbed plaque accumulation results in similar clinical inflammatory responses from
both gingiva and peri-implant mucosa. Future studies of the GCF and plaque
composition will further clarify whether there are any molecular and bacterial differences
at the two experimental mucosal sites. The reversibility of gingivitis and peri-implant
mucositis was demonstrated by the reimplementation of oral hygiene measures over a
second 21-day period. Reinstitution of oral hygiene was not sufficient to restore the
gingival/mucosal conditions observed at baseline (d.0). However, a reduction in the
52
parameters of PlI, GI, AngBS, and GCF volume back to nave condition levels was
achieved at the completion of the resolution phase.
8. References
1.
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
28.
29.
54
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
43.
44.
55
45.
46.
47.
48.
49.
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
56
61.
62.
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
Lamster IB, Hartley LJ, Vogel RI: Development of a biochemical profile for
gingival crevicular fluid. Methodological considerations and evaluation of
collagen-degrading and ground substance-degrading enzyme activity during
experimental gingivitis. Journal of periodontology 1985, 56(11 Suppl):13-21.
Garnick JJ, Pearson R, Harrell D: The evaluation of the periotron. Journal of
periodontology 1979, 50(8):424-426.
Lang NP, Joss A, Orsanic T, Gusberti FA, Siegrist BE: Bleeding on probing. A
predictor for the progression of periodontal disease? Journal of clinical
periodontology 1986, 13(6):590-596.
Lang NP, Wetzel AC, Stich H, Caffesse RG: Histologic probe penetration in
healthy and inflamed peri-implant tissues. Clinical oral implants research
1994, 5(4):191-201.
Jepsen S, Ruhling A, Jepsen K, Ohlenbusch B, Albers HK: Progressive periimplantitis. Incidence and prediction of peri-implant attachment loss.
Clinical oral implants research 1996, 7(2):133-142.
Luterbacher S, Mayfield L, Bragger U, Lang NP: Diagnostic characteristics of
clinical and microbiological tests for monitoring periodontal and periimplant mucosal tissue conditions during supportive periodontal therapy
(SPT). Clinical oral implants research 2000, 11(6):521-529.
Gerber JA, Tan WC, Balmer TE, Salvi GE, Lang NP: Bleeding on probing and
pocket probing depth in relation to probing pressure and mucosal health
around oral implants. Clinical oral implants research 2009, 20(1):75-78.
Van der Weijden GA, Timmerman MF, Nijboer A, Reijerse E, Van der Velden U:
Comparison of different approaches to assess bleeding on probing as
indicators of gingivitis. Journal of clinical periodontology 1994, 21(9):589-594.
Russell AL: A system of classification and scoring for prevalence surveys of
periodontal disease. J Dent Res 1956, 35(3):350-359.
Mombelli A, van Oosten MA, Schurch E, Jr., Land NP: The microbiota
associated with successful or failing osseointegrated titanium implants. Oral
microbiology and immunology 1987, 2(4):145-151.
Silness J, Loe H: PERIODONTAL DISEASE IN PREGNANCY. II.
CORRELATION BETWEEN ORAL HYGIENE AND PERIODONTAL
CONDTION. Acta odontologica Scandinavica 1964, 22:121-135.
Heitz-Mayfield LJ: Peri-implant diseases: diagnosis and risk indicators.
Journal of clinical periodontology 2008, 35(8 Suppl):292-304.
Karoussis IK, Kotsovilis S, Fourmousis I: A comprehensive and critical review
of dental implant prognosis in periodontally compromised partially
edentulous patients. Clinical oral implants research 2007, 18(6):669-679.
Karoussis IK, Salvi GE, Heitz-Mayfield LJ, Bragger U, Hammerle CH, Lang NP:
Long-term implant prognosis in patients with and without a history of
chronic periodontitis: a 10-year prospective cohort study of the ITI Dental
Implant System. Clinical oral implants research 2003, 14(3):329-339.
Schou S, Holmstrup P, Worthington HV, Esposito M: Outcome of implant
therapy in patients with previous tooth loss due to periodontitis. Clinical oral
implants research 2006, 17 Suppl 2:104-123.
57
76.
77.
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
58
90.
91.
92.
59